Acquisition of default inflectional suffix: Japanese adjectivals

Miho Fujiwara
Georgetown University

This study illustrates a case in which a default form of past tense inflectional suffix differs between child language and adult language. Possible accounts for the change of the default form from one suffix to the other would be discussed. The case at hand deals with adjectival inflectional suffixes in Japanese.
Japanese language has two types of adjectivals in one language system (Kuno 1973, Martin 1975). One shows some distributional and/or morphological similarities with nouns (Nominal Adjectives: NAs). The other behaves on a par with verbs (Verbal Adjectives: VAs). These two types of adjectivals are not semantically nor syntactically distinguishable (Urushibara 1993, Nishiyama 1998), but they are morphologically different (Nishiyama 1998). In its predicative use, tense of the clause is expressed differently. NAs are followed by _-da_ (non-past) and _-datta_ (past) and VAs are followed by _-i_ (non-past) and _-katta_ (past).
In the experiment, an elicited task was given to four-year-old children (n=5) and adult control group (n=24). They were asked to say a past form of the adjectival, which they heard its non-past tense form as input. The input consists of two patterns: correct combination of the stem and the non-past tense suffix (VA-i and NA-da) and mismatching combination of the stem and the suffix (VA-da and NA-i). The subjects were also asked to provide the past form of the input, which used novel adjectival stems. These novel stems were made to sound like either VAs or NAs. The novel stems also had matching suffixes (VA.sounding.stem-i and NA.sounding.stem-da) and mismatching suffixes (VA.sounding.stem-da and NA.sounding.stem-i) followed.
The results show that the adults chose an appropriate past form for existing adjectivals even though the stem and the suffix were mismatched. With novel adjectival input, they tended to rely on the non-past suffix in the input to decide a type of adjectivals (VA or NA) to choose the past tense suffix for the stem. However, when the mismatching stem and suffix was given, sometimes they failed to identify a type of the adjectivals. In that case, the NA past tense suffix _-datta_ was chosen as a default.
The four-year-old children produced only the VA past tense suffix _-katta_ most of the time. Even for existing NAs, when they were supposed to use the NA suffix _-datta_, they used the incorrect suffix _-katta_ about 50% of the time. With novel adjectival input, they produced the VA suffix _-katta_ more than 87% of the time regardless of the type of input. The NA suffix _-datta_ was used not more than 8% of the time. These results indicate that the suffix _-katta_ is a default adjectival past form for the four-year-old children.
In Sum, the results of the experiment indicate that the children's default past tense suffix for adjectivals is different from that of adults. This fact raises an interesting questions from learnability point of view: How do children (re-)learn the correct default suffix for adjectivals? What triggers this (re-)learning? In order to answer these questions, this paper discusses the strengths and weaknesses of different possible accounts for this change of the default forms between the children and the adults.


Kuno, Susumu. 1973. The Structure of the Japanese Language. Cambridge: MIT Press.
Martin, Samuel. 1975. A Reference Grammar of Japanese. New Haven: Yale University Press.
Nishiyama, Kunio. 1998. The Morphosyntax and Morphophonology of Japanese Predicates. Ph.D. dissertation. Cornell University.
Urushibara, Saeko. 1993. Syntactic Categories and Extended Projections in Japanese. Ph.D. dissertation. Brandeis University.

< Back to List of Abstracts | Back to PLC23 Home Page >

About the PLC23 Committee
Previously held Penn Linguistics Colloquium: PLC22 (1998), PLC21 (1997)

Penn Department of Linguistics
University of Pennsylvania